Decoder Address reads as 16383

[djconway]

diagram of my wiring that allows me to:
1 - program on the programming track
2 - program on the main
3 - run in DCC mode
4 - run in DC mode

Since I am using center off DPDT switches, there is no shorting between DC-DCC-DCC prog modes.

New to CAD -- Still learning.

Interesting developments, I will need to change the diagram.

Went through the "OK what did you change dummy?" steps in trouble shooting and came up with one big change that I didn't think would change anything.  ==> I moved the command station to the layout room and ran the track A & B, the programming A & B and the loconet cable over to my office.  Not a long run but approximately 30 feet.  This didn't seem to effect some locomotive/decoder combos (BLI/Paragon, Atlas/Lenz) but it raised hell with Atlas/Digitrax and IM/Digitrax.  So in trouble shooting, I came up with an idea that there may be noise-signal degradation or something in the programming lines.  Since the Digitrax PR3 can be used as a standalone programmer I figured I have nothing to loose (except convince) by using the PR3's programming outputs.  Initial results are encouraging, took a DZ123 equipped loco that had been returning all CV's as 255 and tried to read it. Long story short I was able to read ALL CV's as something other than 255, where I expected 35 it read 39(?) or some other random number.  A reset to factory defaults, brought a lot of expected values. A reprogram to what I wanted was successful.
 
I will be  testing this with other locos that could not be read tomorrow.

I knew it couldn't be 10 locos going bad all at once, it had to be some hardware change.
Chnged 2-12-21

[John]

Seems pretty straight forward .. on the off chance that we are missing something .. have you tried disconnecting the setup and using a temp program track just by itself  ..

[djconway]

Seems pretty straight forward .. on the off chance that we are missing something .. have you tried disconnecting the setup and using a temp program track just by itself  ..

I conducted 2 programming tests on this:
Test 1 -- Connect a piece of track directly to the Prog A and B outputs on the DCS200 --- No  change in results.
Test 2 -- Connect the PR3 as a standalone programming module.  -- No change

I am now convinced that I have several 1/2 - 3/4 dead decoders.  They will be replaced, 
As a 23 year user od Digitrax (almost exclusively) its now time to look elsewhere.

[Bangorboy]

One thing to consider:  In the past I learned that the way Digitrax reads decoder CVs is to try each value from 0 thru 255 until it finds a match.  Low values thus read much faster than high ones.  Anyway, I suspect that if so many of the CVs are reading as 255, it could be a false value simply caused by not reading any match at all -- "I tried everything from 0 to 254 and didnt get a response, so it must be 255."  Same applies to the address -- The highest possible value is 16,383, so if it tried every value from 0 to max and found no response, it would end at 16,383.
I think your decoder didn't change values.  You simply have a problem between the device reading it and the decoder.  There's an open or high resistance circuit -- bad connectiion, dirty track, loose contact, etc. -- somewhere.
// Posted this before i noticed I missed a whole page of replies.  I'll read through them now.  Hope I didn't step on anybody.//Bill

[mmyers]

I would try writing a new address to the decoder and try it out. I know CV8 doesn't seem to be taking the write but I have been doing it since you too and still try the old fashioned way. My thinking is why read it if I will need to write the correct value anyway. Started with a Big Boy and DN93's so that was the only option available.

Martin Myers

[djconway]

Been doing a bit more testing - thinking - cussing.
I am beginning to think that my control station may be too far from my test bench.
I have the layout in another room with the rail and program wires (about 30 feet 16gauge copper wire) running to my "office" test loop.
Going to try one of the locomotives that run but will not program directly to the command station program outputs.

Who knows, it may work.

[John]

Been doing a bit more testing - thinking - cussing.
I am beginning to think that my control station may be too far from my test bench.
I have the layout in another room with the rail and program wires (about 30 feet 16gauge copper wire) running to my "office" test loop.
Going to try one of the locomotives that run but will not program directly to the command station program outputs.

Who knows, it may work.

Waiting for results .. this is new information

[Aleuck]

I have run into this problem, I changed the program mode ie: paged direct etc.. then I was able to read back correctly.

[djconway]

I have run into this problem, I changed the program mode ie: paged direct etc.. then I was able to read back correctly.

Going to have to try this.

Moving the command station connection had no effect.

Thanks Aleuck -- I steped through all programming modes and the results were the same. 

Its time for another help request to Digitrax -- I know its not a dirty track / wheels / pickup issue, if the loco runs fine on DC.

Thanks for the ideas folks but this reeds like a hardware issue in the decoder.

[djconway]

Made some progress on this today.

1 - Switched the 12v power supply for the PR3 to a 15v
2 - Reconfigured JMRI to have the PR3 function as a stand alone programmer
3 - Re-try one of the confused locos.

When I tried to read one of the confused loco, not the number I expected but not 255 (progress?)
Try resetting the decoder, address reads as 3 (this is what I expected, yes progress) and the loco runs
Try setting address to loco number, address reads as expected, and the loco runs
Set speed table, acceleration, deceleration and all the other cv's -- so far this seems to be working

So this brings up a few things that I'm going to have test.
1 - Was my 12v power supply not functioning properly?
2 - Is there some sort of signal interference caused by having the command station remote from my test and setup ares? (Approximately 30ft)
3 - is my command station dieing?

[mu26aeh]

1 - Switched the 12v power supply for the PR3 to a 15v

I have a friend who models in O scale that was having odd issues with DCC and such.  Everything he tried, nothing fixed the problem.  Only thing he hadn't done was adjust his power supply.  I can't remember the exact numbers, but he upped his Volts and it cleared all his problems.

[John]

Made some progress on this today.

1 - Switched the 12v power supply for the PR3 to a 15v
2 - Reconfigured JMRI to have the PR3 function as a stand alone programmer
3 - Re-try one of the confused locos.

When I tried to read one of the confused loco, not the number I expected but not 255 (progress?)
Try resetting the decoder, address reads as 3 (this is what I expected, yes progress) and the loco runs
Try setting address to loco number, address reads as expected, and the loco runs
Set speed table, acceleration, deceleration and all the other cv's -- so far this seems to be working

So this brings up a few things that I'm going to have test.
1 - Was my 12v power supply not functioning properly?
2 - Is there some sort of signal interference caused by having the command station remote from my test and setup ares? (Approximately 30ft)
3 - is my command station dieing?

I think you might be on to something -- the PR3 manual says 12-15V .. but recomends the PS14 ..     Current Rating   14V DC 300ma ==

 Glad to see you are making progress ..  Maybe measure your power supply to verify the theory .. It's possible the 30 feet of wire to the program track is causing enough signal loss -- which may not be enough to get accurate readings -- I suppose you could move the command station right next to the programming location to test if anything changes ..

[peteski]

The big question is why this setup worked before (before he put those engines away few years back). Or was the programming hardware setup changed since the last time those engines were programmed?

[djconway]

The big question is why this setup worked before (before he put those engines away few years back). Or was the programming hardware setup changed since the last time those engines were programmed?

Around the end of December I moved the command station into the layout room.  That setup worked for some locos, what really confused me was when I started tuning up my IM F3 and F7's to match them up with BLI units. The first 4 no problem then 16383, pulled a few more out of the box and some worked some didn't.  Thinking I may have had a problem with moisture or static or ???!!  All of the F units had the same DN163I1 family of decoders, so I was thinking it was a problem with the decoders.  Well I pulled out a few more locos with different types of decoders (all Digitrax, I know how you love Digitrax ) seemed to have random problems. Finally after a few weeks of head scratching and cussing I finally sat down and said "OK what is the common thread and what changed?  I didn't think moving the command station was a factor, after all I used 16 gauge wire for the run from my office test bench, to the command station, power to run locos, detect shorts seems fine.  I should have started my trouble shooting with "what changed?"  There may yet be some issues with my 18 year old Super Chief system.  Going to have to borrow a O-scope and check things out more thoroughly. 

[peteski]

I  troubleshoot computer equipment for a living, and "what's changed since it worked last" is one of the first questions that is asked.  But my customers often take me on unintentional slide trips or dead ends (or say "nothing's changed", until they find out that someone else messed with the setup without telling them about it).  I also took an "analytical troubleshooting" course back in the '80s and we had to troubleshoot a problem with a fantasy square-donut-making-machine.  Fun stuff!